Data Analysis: an AI-era textbook project

Cambridge University Press Open Access discussion

Gábor Békés (CEU)

2025-10-22

This presentation

This project is part of Gabors Data Analysis and AI Lab

gabors-data-analysis.com/lab/

About me and this slideshow

About me

  • Gábor Békés, Associate Professor of Economics, Central European University (CEU)
  • Co-author: Data Analysis for Business, Economics, and Policy (Cambridge University Press, 2021)

About this slideshow

  • Showcase my proposal for an AI era textbook, based on open access materials and collaboration with tech.

  • Think about opportunities with university-publisher-tech partnerships in the AI era

What we have now

Unchanged mission

Democratizing access to data analysis knowledge

Past

  • Language of the book
  • Case studies: global, accessible, real life
  • Code free, coding courses

. . .

Present

  • This project: textbook + classroom innovation, global reach

In the works

  • Textbook 2nd edition
    • Error correction / minor improvements (many…)
    • Small new bits based on feedback (a dozen)
    • AI boxes (few/chapter)

Use of Artificial Intelligence in classes

Use of Artificial Intelligence in classes

“what is the share of students already using AI in the coursework”

The AI-era textbook challenge

The AI-era textbook challenge: Textbook vs AI

🧑 Textbooks = transform world knowledge

  • Distill, curate what matters
  • Translate to easy understanding
  • Provide context for students
  • Offer examples and case studies

💻 GenAI does all this

  • In fact it excels all of this

The AI-era textbook challenge: content consumption

  • GenAI ability to summarize, translate and give examples changes how we interact with information
  • People consume information through AI agents

    • Google search –> Gemini summary
    • Ask ChatGPT etc directly
    • Business model based on discovery challenged
    • Search traffic to educational content halved (says the AI)

The AI-era textbook challenge: Why still textbooks

  • GenAI
    • built on all of the internet. Hence errors, imprecision
    • is a stochastic parrot. It’s unstable
  • Textbook
    • Curated and supervised content – precision, details, accuracy
    • Stable and reliable (no variation in output) learn –> exam
    • Faculty trust

:::

Stochastic Parrot

GenAI hallucinates 3-15% of factual claims depending on domain specificity

The AI-era textbook challenge: options

  • Gated
    • Book is print + ebook, not available online,
    • not fed to AI
    • Clear business model, but challenged by AI
  • Open and integrated
    • textbooks are online first (gated or not)
    • AI assistant built in (but prevented to use it as training data)
    • Technology direction clear, business model is not

The AI-era textbook pilot

The AI-era textbook challenge: the pilot

💡 Innovative overhaul of what a textbook is

  • This is a pilot project for the open and integrated model
  • Built a coalition of Github, Inc, CEU, in talks with others
  • Hopefully Cambridge University Press joins

Discuss Open Access

  • Fee
  • Collaboration
  • Partnership: content expansion + innovative edu-tech + market reach

Demo !

Sneak preview: demo

Online version of chapter 10 (freely available) + Integrated tools

gabors-data-analysis.com/book

Features: Interactive code, AI assistant, embedded dashboards, version control integration

Open Access

Github, Inc as parner

  • My Lab is building a collaboration with Github, Inc
    • SF based software dev, owned by Microsoft, 300m users
    • Open source ethos core to mission
    • Fanbase
  • I joined Microsoft AI for Good Lab as advising Fellow (minor role)
    • Experimentation in AI in education

OA

  • Github pays a single fee to help make an open access version
  • Helps with AI-era textbook development
  • Asks we use GH/MS products on tech bits (dashboard cloud, code)
  • Joins promo
  • GitHub’s community (100M+ developers) as natural audience, opening to social sciences

:::

Experimentation with a new model

Sales and revenue

  • Revenue model: OA fee + sustained print/ebook sales from enhanced visibility*
  • Clear alternative to purchase (albeit no offline version)
  • Massive PR around website
    • Greater college adoption
    • Professionals buying books as market potential after Github vibe

Experimentation and PR

  • Innovative project, PR from all parties
  • Being in the conversation vs defensive strategies
  • Early-mover position in AI-integrated academic publishing*

:::

What’s next

Development

  • From pilot to proper textbook site
  • Extending dashboards
  • New case studies
  • New courses (Data Analysis with AI already)
  • More AI integration
  • Lab’s main project: test how AI works in data analysis education

Thank you

Questions & Discussion

Contact: bekesg@ceu.edu

Project: gabors-data-analysis.com/lab/